查看文章

Driving in stop-&-go (S&G) traffic can be a frustrating experience and can lead to an increase in fuel emissions, driver distress and safety risks. Past studies have modelled the time spent in S&G traffic as one of the factors for the disutility of a route. These studies did not account for the effect of the number of S&G waves, which has been found to bear a closer relationship with driver discomfort. This thesis tests the research hypothesis that an increase in the number of S&Gs on a route increases its disutility for a driver. As a proof of concept, a Stated Choice (SC) experiment was initially conducted on a sample of university staff and students. A D-efficient pivot design technique was used to generate the set of choice tasks. The competing routes were defined in terms of the number of S&Gs along with other travel attributes. The collected data was analysed using a Random Parameter Error Component Logit model to account for the taste heterogeneity and serial correlation among the choice tasks. The results showed that the number of S&Gs negatively impacted the utility of a route, thus validating the proposed hypothesis. Another follow-up SC study was then conducted on a sample of general commuters in Sydney. The data from this study was analysed using a Latent Class Choice Model which relaxed the distributional assumption associated with the previous statistical model. The obtained results showed a negative effect for the number of S&Gs, except for nearly one quarter of the sample which was indifferent towards this attribute. Lastly, a driving simulator experiment was conducted to further understand the role of S&G traffic characteristics on …